def main():
solver = DifferentialEvolutionSolver(ND, NP)
solver.SetRandomInitialPoints(min=[-100.0] * ND, max=[100.0] * ND)
solver.SetEvaluationLimits(generations=MAX_GENERATIONS)
solver.enable_signal_handler()
strategy = Best1Bin
stepmon = VerboseMonitor(1)
solver.SetGenerationMonitor(stepmon)
solver.Solve(wavy, ChangeOverGeneration(generations=50), \
strategy=strategy, CrossProbability=1.0, ScalingFactor=0.9, \
sigint_callback = plot_solution)
solution = solver.Solution()
return solution, solver
def test2(monitor, diffenv=None):
if diffenv == True:
#from mystic.solvers import DifferentialEvolutionSolver as DE
from mystic.solvers import DifferentialEvolutionSolver2 as DE
elif diffenv == False:
from mystic.solvers import NelderMeadSimplexSolver as noDE
else:
from mystic.solvers import PowellDirectionalSolver as noDE
from mystic.termination import ChangeOverGeneration as COG
from mystic.tools import getch, random_seed
random_seed(123)
lb = [-100, -100, -100]
ub = [1000, 1000, 1000]
ndim = len(lb)
npop = 5
maxiter = 10
maxfun = 1e+6
convergence_tol = 1e-10
ngen = 100
crossover = 0.9
percent_change = 0.9
def cost(x):
ax, bx, c = x
return (ax)**2 - bx + c
if diffenv == True:
solver = DE(ndim, npop)
else:
solver = noDE(ndim)
solver.SetRandomInitialPoints(min=lb, max=ub)
solver.SetStrictRanges(min=lb, max=ub)
solver.SetEvaluationLimits(maxiter, maxfun)
solver.SetEvaluationMonitor(monitor)
#solver.SetGenerationMonitor(monitor)
tol = convergence_tol
solver.Solve(cost, termination=COG(tol, ngen))
solved = solver.Solution()
monitor.info("solved: %s" % solved)
func_max = -solver.bestEnergy
return solved, func_max
def main():
from mystic.solvers import DifferentialEvolutionSolver2 as DifferentialEvolutionSolver
solver = DifferentialEvolutionSolver(ND, NP)
solver.SetRandomInitialPoints(min=[-100.0] * ND, max=[100.0] * ND)
solver.SetEvaluationLimits(generations=MAX_GENERATIONS)
solver.SetGenerationMonitor(VerboseMonitor(30))
solver.enable_signal_handler()
strategy = Best1Exp
#strategy = Best1Bin
solver.Solve(ChebyshevCost, termination=VTR(0.01), strategy=strategy, \
CrossProbability=1.0, ScalingFactor=0.9, \
sigint_callback=plot_solution)
solution = solver.Solution()
return solution
def de_solve(CF, a4=None, a5=None):
"""solve with DE for given cost funciton; fix a4 and/or a5 if provided"""
minrange = [0, 100, 100, 1, 1]
maxrange = [100, 2000, 200, 200, 200]
interval = 10
if a5 != None:
minrange[4] = maxrange[4] = a5
interval = 20
if a4 != None:
minrange[3] = maxrange[3] = a4
if interval == 20: interval = 1000
solver = DifferentialEvolutionSolver(ND, NP)
solver.enable_signal_handler()
stepmon = MyMonitor(interval)
solver.SetRandomInitialPoints(min=minrange,max=maxrange)
solver.SetStrictRanges(min=minrange, max=maxrange)
solver.SetEvaluationLimits(generations=MAX_GENERATIONS)
solver.SetGenerationMonitor(stepmon)
solver.Solve(CF,termination=ChangeOverGeneration(generations=50))
solution = solver.Solution()
return solution, stepmon
# DEsolver inputs
MAX_GENERATIONS = 2000
ND, NP = 3, 30 # dimension, population size
minrange = [0., 0., 0.]
maxrange = [50., 50., 10.]
# prepare DESolver
from mystic.solvers import DifferentialEvolutionSolver2 \
as DifferentialEvolutionSolver
from mystic.termination import ChangeOverGeneration, VTR
solver = DifferentialEvolutionSolver(ND, NP)
solver.enable_signal_handler()
solver.SetRandomInitialPoints(min=minrange,max=maxrange)
solver.SetEvaluationLimits(generations=MAX_GENERATIONS)
solver.Solve(cost, termination=ChangeOverGeneration(generations=100))
if __name__ == '__main__':
# x0, y0, R0
#guess = [1,1,1] # bad initial guess
#guess = [5,5,1] # ok guess
guess = [10,15,5] # good initial guess
# plot training set & training set boundary
pylab.plot(xy[:,0],xy[:,1],'k+',markersize=6)
c = circle(x0, y0, R0)
pylab.plot(c[:,0],c[:,1],'r-',linewidth=2)
legend = ['random points','generating circle : %f' % R0]
pylab.axis('equal')
